Master Thesis Development of Retrieval-Augmented Generation, Pipeline Components

Summary

Join Bosch Rexroth and contribute to a Master's thesis in one of three areas: developing an ML model for parsing industry documents for an RAG pipeline; developing custom parsing and chunking tools for document ingestion into a vector database; or training/fine-tuning a custom LLM model for smart hydraulic services. The position requires a Master's degree in Computer Science or a related field, basic Linux and Python scripting skills, and foundational knowledge in at least one area such as Machine Learning, NLP, LLMs, RAG, information retrieval, vector databases, or OCR. Familiarity with CLI tools is preferred. The role is primarily remote within Germany, with monthly on-site visits required. The position offers flexible work arrangements and is open to candidates who prefer working from home or at the Bosch location in Lohr am Main. The thesis duration is 3-6 months, and university enrollment is a requirement. Bosch Rexroth values diversity and inclusion.

Requirements

• Education: Master studies in the field of Computer Science, Information Technology or a comparable degree program in Computer Science
• Experience and Knowledge: basic in Linux; hands-on experience with Python scripting; basic knowledge (can be coursework, personal projects or practical exposure) in at least one of the following: Machine Learning, Natural Language Processing (NLP), Large Language Models (LLMs), Retrieval-Augmented Generation (RAG) concepts, information retrieval, vector databases or Optical Character Recognition (OCR)
• Personality and Working Practice: you are an independent and autonomous individual with a strong ability to learn and a willingness to familiarize yourself with new subject areas
• Good in German

Responsibilities

• Develop and train an ML Model to parse industry-specific documents as entrypoint for an RAG pipeline
• Develop custom parsing and chunking tools for industry-specific documents and their ingestion into a vector database
• Train and/or fine-tune a custom LLM model for use-case specific usages within the scope of smart hydraulic services

Preferred Qualifications

Familiarity with CLI tools is preferred

Benefits

• The team mainly works remotely, with mobile working within Germany as the standard mode. Monthly visits to the common team on-site days are required
• You want to work flexibly from your home in Germany or prefer working at the Bosch location in Lohr am Main? For positions with the addition “remote possible”, you can agree on the appropriate collaboration for your task together with your manager and your team within the framework of Smart Work